Television viewers have more options than ever before when it comes to when and how they enjoy their favorite programs. In addition to traditional “real-time” viewing and time-shifting using a VCR or DVR, people with broadband internet access have the option of viewing on the internet. Some programs are even being streamed to mobile phones. All of these innovations have led advertisers and agencies to question whether the old rules of engagement still apply in this new age of consumer control, even though evidence presented to date suggests that the impact of DVRs has been greatly overstated. Millward Brown’s recently released CTV study provides further evidence that growing consumer control over when and how content is viewed does not spell the death of TV advertising as we know it.
CTV is a shorthand reference developed to describe video content viewed under conditions of the viewer’s choosing—in other words, whenever and wherever they choose. The concept spans any type of video content or channel, but our initial CTV study explored the effects of a 30-second ad shown during a prime-time network program on three different platforms: real-time TV, DVR via time-shifting, and on a network Web site. The study used ads for leading quick service restaurants, consumer packaged goods, and financial services brands.
The CTV study included more than 3,000 respondents who were pre-recruited to watch a prime-time television show in one of the three viewing environments. Then, within 48 hours of viewing, participants answered an online survey. The study was conducted in conjunction with the launch of new ad campaigns, with control cells used to gauge the level of background awareness.
First let’s review some results for the “normal” viewing group. What did this study tells us about the effectiveness of traditional TV viewing in the United States?
Traditional TV Viewing
Only 30 percent of respondents in the live viewing group claimed to pay attention to the ads during the commercial breaks. This did not, however, prevent 59 percent of them from recognizing the ad when shown it in the survey (compared to a control cell score of 22 percent). So—a higher proportion of people recognized one of the commercials than actually claimed to have attended to the ads! That, of course, is the beauty of TV. Even though people claimed to ignore the ads, (as 29% did), talk to someone else in the room (24%), or read e-mail while the ads aired (10%), they were still exposed to the ads. Only the 18 percent that claimed to have left the room really avoided ad exposure.
DVR Viewing
In the DVR viewing cell, only 20 percent of respondents claimed to have paid attention to the ads during the test program; 60 percent claimed to have skipped over them. However, ad recognition, at 50 percent, was not substantially lower than it was among traditional TV viewers (59%). Brand-linked ad awareness (which was asked before ad recognition in the survey, using only the brand name as a prompt) was actually identical between the two groups. This result confirms the findings I reported last year in this post, and demonstrates that the act of fast-forwarding does not prevent ad exposure. It is notable, however, that DVR owners were less likely to claim to enjoy watching the test ad (52% versus 75%) and more likely to say it was hard to tell what was going on in the ad (18% versus 8%).
Online Viewing
Ad- and brand-related measures were highest (by a significant margin) among the online viewing group. More than half of this group (51%) reported paying attention to the ads during commercial breaks. This high degree of attention resulted in brand-linked ad awareness of 82 percent and specific ad recognition of 83 percent. We believe that the higher levels of response are due to two different factors:
- Online media consumption is typically an active, “lean-forward” experience. People are used to focusing on web-based content and controlling the way they attend to it. This heightened level of attention may carry over to TV programs viewed online, causing viewers to watch them more intently than they would on regular TV.
- Online viewers had the opportunity to see the test ad multiple times during the course of the program. Frequency does count, particularly when those exposures happen within a relatively short time frame. While the multiple exposures (3 to 6) ensured that people did remember seeing the ad, a possible downside to frequency was also apparent. People in this cell were more likely to report being tired of seeing the test ads (27%, versus 15% among the regular TV viewing cell).
The observations of TV viewing behavior offered by this study are interesting. But I hope that the results do not further encourage the U.S. marketing community to pay undue amounts of attention to DVR owners and online TV viewers. In the overall scheme of things, these are still two relatively small groups of people. For all the hype about YouTube, pre-rolls and iPods, “regular” TV viewing still accounts for the vast majority of time spent with video in the United States. Assuming some overlap between DVR-owning households and those who view online, we’d estimate that about two-thirds of U.S. households still view TV the old-fashioned way, and perhaps three-quarters of all TV viewing time still occurs in a passive, “lean-back” mode.
I believe that a few statistics will support these assertions. TV ownership in the United States is nearly universal, with 98 percent of households owning at least one set (source: TV Dimensions). DVR ownership, on the other hand, has only reached 13 percent, making eMarketer’s prediction that 21 percent of households will use DVRs in 2007 seem a bit of a stretch. And only about a third (36%) of DVR users in the DVR cell claimed to time-shift more than half of their viewing.
Online video viewing is also gaining ground, but its penetration is limited and intermittent. While the ability to view online video is available in 50 million homes with broadband internet access, a survey conducted last fall by Nielsen/NetRatings reported that only 11.4 percent of adults reported viewing online video in the past 30 days. Harris Interactive reported that 41 percent of households with online access reported viewing videos online at network sites sometime in 2006. Based on these figures, my rough calculations suggest that 27 percent of the U.S. population might be expected to view full-length TV programming online at some time in 2007. However, the majority of online users thus far have reported using it as a means to catch up with missed viewing, so watching TV episodes online is probably a fairly infrequent behavior for most people.
Overall, the results of the CTV project offer an interesting glimpse of what the future holds for TV advertising. Based on the facts presented here, that future looks less bleak than we might have expected. My conclusion is that the combination of moving pictures and sound remains an incredibly powerful communication medium, irrespective of which device is used to view it. The findings from the DVR cell suggest that advertisers do need to find a way to ensure that fast-forwarded content still makes sense to the viewer, or that it at least triggers memories of an ad viewed previously under normal viewing conditions. The results from the online cell suggest that advertisers need to be more savvy about ad rotation online. Use of multiple executions in one program is likely to be more effective, and less irritating, than repeated exposure to the same ad.
So what do you think? Am I too optimistic about TV’s future, online and offline? What else do we need to take into account when we try to examine that future?
Nigel Hollis Millward Brown future of TV online video TV viewing ad awareness DVR PVR CTV
(28 votes, average: 3.25 out of 5)
Loading ...
Email This Post
March 29th, 2007 at 12:18 pm
Here is a faulty assumption that doesn’t get discussed often. We assume that all creative is equal when we purchase units. We know that’s not the case. If memory serves me correctly more enjoyable ads have a magnifying effect on all measures. Basically all this hand-wringing about the death of the 30-sec ad is stupid. Maybe we should modify it to be the death of the mediocore 30 sec- ad.
Now have you done a study of how great creative gets affected by DVR or CTV or whatever term we want use. Have Cannes winning creative or One Show creative and see what the numbers say then, vs. just ordinary ads–I’m sure we’ll get interesting results.
Until the numbers are associated with quality of creative, then their ability to tell you anything is limited
April 2nd, 2007 at 8:01 am
Thanks for the comment, rukallstar. If you are referring to many people in the media world, you may be correct in saying that people assume all creative is equal. Based on our extensive experience of tracking the in-market impact of advertising on awareness, attitudes and sales I would hope no one at Millward Brown would hold that belief! You are correct that copy varies dramatically in its ability to turn exposure into a positive influence on the brand. And yes, people have always tried to avoid (and more importantly forgotten) mediocre ads.
This experiment was designed to highlight the impact of viewing conditions on likely ad effect, so holding the creative constant was important. I may be able to find out what the differential effects were across the ads tested but I would expect their effects to hold constant relative to each other across the different viewing conditions.
If I can find out more I will let you know.
Nigel
April 2nd, 2007 at 3:53 pm
POSTED ON BEHALF OF ERIK DUPLESSIS
Hi Nigel
The CTV study is great and timely, especially the bit about DVR owners saying that they don’t watch television.
When I bought my DVR I found that I could time shift programs (Desperate Housewives) to a more convenient time. I also discovered that I could watch it faster as I can fast-forward through the break. I ALSO DISCOVERED THAT ONE NEEDS TO CONCENTRATE ON THE ADS TO KNOW WHEN TO STOP FAST-FORWARDING.
If you asked me I would have told you that I use my DVR to avoid advertisements, when in fact my DVR is making me pay attention to the advertisements.
To test whether or not people see ads when they give attention to a fast-forwarded ad-break, we recorded one at triple speed and showed it in a cinema. We asked the respondents to name the ads they saw, to tell us if they have seen them before, and to give each points out of ten in terms of liking.
The lowest recall was 40% and the two highest were 90%, remember this is not prompted recall. This is way above what ads in a normal television break will achieve.
It is clear that when the ad was seen before it did better in the experiment - very few people claimed that they were seeing the ad for the first time.
Much like CTV found for enjoyment, we found that the liking scores were lower for the fast-forwarded advertisements. However, there was a very strong correlation with the liking scores we achieved on Adtrack for the same ads. Presumably this can only be because the emotional reaction (liking) was established by the previous exposures.
Most of the research I have read on DVRs makes the assumption that ads which are fast-forwarded are totally ineffective. I believe that they are more effective because they get attention (even if the viewer believes he is not giving attention).
Well done.
Best regards
ERIK
April 3rd, 2007 at 8:06 am
So maybe the answer is for ads on shows that are knownn “DVR” favourites to be made in such a way that when you fast forward over them you still see the brands/concepts…less variability in the ad means you will see more when you FF over it..hey back to “billboard” ads with a voice over !
April 3rd, 2007 at 1:50 pm
Thanks for the comments Erik and Andrew.
Erik, what you say is in agreement with a lot of research into the effect of DVRs (rather than pundits opinions!). People attend to DVRs when fast forwarding in a way that is similar to navigating online content - both are “lean forward” activities.
Andrew, as Erik suggests, I think the key is to make sure that people see the ads in normal viewing mode at least once before fast forward exposure by placing them in time sensitive content, e.g. live coverage of sports events. Failing that I would hope that we could be a bit more creative than simply creating an ad banner with sound. It would, however, place a premium on creating ads with iconic imagery, e.g. Burger King’s King character, Guinness’s beer in glass shot or the Geico Gecko.
April 3rd, 2007 at 3:11 pm
True - but I don’t think you can’t rely on people seeing ads in full at least once. I rarely watch TV when it is “real time” simply to make sure I can skip the ads. Creating ads that are a mixture of “static” iconic imagery and more dynamic content gets the best of both worlds, and there is the hope that the “static” imagery may trigger a desire to watch the ad in real time….
April 4th, 2007 at 7:19 am
I can tell you are a hard core ad avoider, Andrew. Luckily for advertisers you are in a minority. Recorded content makes up more than 50% of viewing for only 1 in 3 people today and I confidently predict that this will drop as DVRs penetrate a broader market. In my experience it is type A ad execs, marketers, and techies who are most likely to have a DVR. Most people are content to go with the flow.
Of course, to your point, that does pose a problem for brands that need to reach those type As, and in which case your solution is one way to do so.
April 4th, 2007 at 4:28 pm
I’ve read some of the coverage this study has gotten online (on Clickz, Adweek, etc.) and I’m concerned about the amount of emphasis given to the higher levels of awareness and impact from the online advertising. The stories seem to suggest that people are really excited about watching advertising online!
While I would suggest that the enhanced level of attention may have resulted from people viewing programs in what is currently an unfamiliar viewing environment. I mean, these people think they’re supposed to be watching a TV program. That’s what they were recruited to do, right? So, they’re watching a program on this tiny little laptop screen, and commercials come on. Maybe they watched the commercials because they wanted to be sure they were right there when the program started up again. They wanted to be conscientious, and they had no way of knowing if there were going to be 3 to 5 minutes worth of commercials, as there would be on TV, or if there would just be one short ad.
So, I think that when people get used to encountering ads in online video content, and they catch on to the rhythm of the ad breaks, they will treat those ad breaks with the same amount of respect that they give to TV ad breaks.
April 6th, 2007 at 7:07 am
I wonder how the recall numbers across media would have panned out if the ad had been shown only once in the program on the web site, or shown 3-6 times on TV with the other commercials remaining the same across both.
Would have been interesting to see the findings, and then drawn a conclusion on whether supplementing with online spends is really worthwhile.
PS : Loved Erik’s experiment - very practical & insightful !
April 6th, 2007 at 7:37 am
Good question, and the answer is that we just don’t know for sure. My guess would be that even if we ensured the same number of exposures, online would win because of the lean forward context.
April 24th, 2007 at 7:46 am
Nigel I loved the post. Thanks. In my mind the biggest difference between the old format and the new is that previously people had no choice but to sit through ads. And as a result advertising was built around this constraint. Now that the constraint has been lifted the way we advertise needs to change. As a consumer I want to only see a high-level description of a product or service and then based on how well it fits my needs I want to drill down and explore more. This is one reason why Google text advertising works so well. Obviously the percentage of consumers whose viewing habits have been altered still remains small. But I believe it’s inevitable.
Sites like NBC force views to see the entire ad when watching online video. This is a bad idea and will hurt the network in the long run. On the other hand, sites like MSNBC place banner ads next to the video. This seems to be a more effective strategy. Those are my 2 bits.
April 24th, 2007 at 8:46 am
Hi Rishi, I am glad you liked it.
Interestingly I saw a presentation in Paris last week that reminded me how much people enjoy watching regular TV. Admittedly the study was conducted by Thinkbox the television marketing body for all the UK commercial broadcasters but it did seem to be pretty unbiased. 20 households’ viewing was monitored and then classified. Only about 10% of people were classified as active ad avoiders. Most ads were viewed with at least partial engagement and some of the footage of people discussing ads and singing along with them was pretty compelling. Again, it reminds me that the future of interactive TV will be limited to some viewers, some of the time. Most people are just happy to watch what gets served up and it will take a decade at least for significant behavior change to set in.